DeepSeek R1 — lokalny model AI od DeepSeek
Autor: Jakub Rusinowski · Ostatnia aktualizacja:
Najnowsza sensacja świata open-source. DeepSeek R1 rywalizuje z topowymi modelami własnościowymi w zadaniach rozumowania i kodowania, używając architektury Mixture-of-Experts. Znany z wyjątkowych zdolności logicznych i matematycznych.
Warianty
Najmniejszy wariant DeepSeek R1 potrzebuje około 6 GB VRAM przy Q4_K_M — skwantyzowane wagi plus narzut frameworka, bez pamięci podręcznej KV.
| Model | VRAM przy Q4 | VRAM | Kontekst | Uruchom |
|---|---|---|---|---|
| DeepSeek R1 Distill Llama 8B → 8B | ~5.6 GB | 128,000 | ollama run deepseek-r1:8b | |
| DeepSeek R1 Distill Qwen 32B → 32B | ~20.1 GB | 128,000 | ollama run deepseek-r1:32b | |
| DeepSeek R1 Distill Qwen 14B → 14B | ~9.3 GB | 128,000 | ollama run deepseek-r1:14b | |
| DeepSeek R1 (671B) → 671B | ~405.9 GB | 128,000 | ollama run deepseek-r1:671b |
Pamięć to skwantyzowane wagi plus narzut przy Q4_K_M, liczone tym samym silnikiem co sprawdzarka GPU i VRAM.
Jak uruchomić DeepSeek R1 lokalnie
Zainstaluj Ollama, a potem pobierz tag.
ollama run deepseek-r1:8bWybierz rozmiar powyżej, aby zobaczyć jego VRAM, szacowaną szybkość i polecenie instalacji.
Licencja
Commercial use permitted. No usage restrictions beyond attribution.
Dotyczy: DeepSeek R1 Distill Llama 8B, DeepSeek R1 Distill Qwen 32B, DeepSeek R1 Distill Qwen 14B, DeepSeek R1 (671B)Polecane GPU
Najtańsze GPU z naszego katalogu, które uruchomi DeepSeek R1 lokalnie (min. 6 GB VRAM), to Intel Arc B570 (10 GB).
Czy uruchomię DeepSeek R1 na moim GPU?
- DeepSeek-R1-Distill-Qwen-14B na AMD Radeon RX 7800 XT
- DeepSeek-R1-Distill-Qwen-14B na AMD Radeon RX 7900 XT
- DeepSeek-R1-Distill-Qwen-32B na AMD Radeon RX 7900 XTX
- DeepSeek-R1-Distill-Qwen-14B na AMD Radeon RX 9060 XT 16GB
- DeepSeek-R1-Distill-Llama-8B na AMD Radeon RX 9060 XT 8GB
- DeepSeek-R1-Distill-Qwen-14B na AMD Radeon RX 9070
- DeepSeek-R1-Distill-Qwen-14B na AMD Radeon RX 9070 XT
- DeepSeek-R1-Distill-Qwen-14B na Apple M1
- DeepSeek-R1-Distill-Qwen-32B na Apple M1 Pro
- DeepSeek-R1-Distill-Qwen-32B na Apple M2 Pro
- DeepSeek-R1-Distill-Qwen-32B na Apple M3 Pro
- DeepSeek-R1-Distill-Qwen-32B na Apple M4
- DeepSeek-R1-Distill-Qwen-32B na Apple M5
- DeepSeek-R1-Distill-Llama-8B na Intel Arc B570
- DeepSeek-R1-Distill-Qwen-14B na Intel Arc B580
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 3060 (12GB)
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 3070
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 3070 Ti
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 3080 (10GB)
- DeepSeek-R1-Distill-Qwen-32B na NVIDIA GeForce RTX 3090
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 4060
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4060 Ti 16GB
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4070
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4070 Super
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4070 Ti
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4070 Ti Super
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4080
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 4080 Super
- DeepSeek-R1-Distill-Qwen-32B na NVIDIA GeForce RTX 4090
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 5060
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 5060 Ti 16GB
- DeepSeek-R1-Distill-Llama-8B na NVIDIA GeForce RTX 5060 Ti 8GB
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 5070
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 5070 Ti
- DeepSeek-R1-Distill-Qwen-14B na NVIDIA GeForce RTX 5080
- DeepSeek-R1-Distill-Qwen-32B na NVIDIA GeForce RTX 5090
- DeepSeek-R1-Distill-Qwen-32B na NVIDIA L40S
- DeepSeek-R1-Distill-Qwen-32B na NVIDIA RTX 6000 Ada Generation
DeepSeek R1 — najczęściej zadawane pytania
How much VRAM does DeepSeek R1 need?
DeepSeek R1 needs about 6 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: DeepSeek R1 Distill Llama 8B (6 GB, Q4_K_M); DeepSeek R1 Distill Qwen 32B (20 GB, Q4_K_M); DeepSeek R1 Distill Qwen 14B (9 GB, Q4_K_M); DeepSeek R1 (671B) (406 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run DeepSeek R1 on an RTX 4090 (24 GB)?
Yes — DeepSeek R1 runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for DeepSeek R1?
Q4_K_M is the best balance of quality and VRAM for DeepSeek R1 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run DeepSeek R1 with Ollama?
Install Ollama, then run: ollama run deepseek-r1:8b. This downloads DeepSeek R1 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.